404 Valerie Fridland
scales, listeners may not have used “Northern” versus “Southern” dialect cri-
teria at all, but were perhaps instead using only intra-Southern criteria com-
paring more rural vs. less rural or more educated vs. less educated sounding
tokens. So, when rating tokens with greater degrees of Southern shift, the
differences Memphians believe to exist between themselves and inhabitants
of other states like Mississippi and Arkansas may be coming into play.
To gain more insight into this contradictory behavior, a ¿ nal perception
study was designed to examine whether intra-regional norms associated with
urban vs. rural speech played into Memphians’ ratings of synthesized vowel
tokens (Fridland and Bartlett 2007). For this study, Memphians were given the
same synthesized vowel token perception test used in the previous perception
study. This time, instead of rating “Southernness,” listeners were instructed
to rate the relative ruralness of Southern shifted and non-shifted tokens, and,
in a separate test, how educated and pleasant the tokens sounded. Participants
were considered accurate when they selected the token shifted most toward
Southern norms in each token-set as the most rural member of the pair, allow-
ing comparison of ruralness accuracy to Southernness accuracy from the ear-
lier perception study.
Before beginning the test, participants were asked to ¿ ll out a brief demo-
graphic questionnaire and each participant was asked to de¿ ne the concept
“rural.” In order to ensure we knew what participants were responding to
conceptually, the analysis was based on participants’ ideas of rurality rather
than a set de¿ nition which may or may not have matched with that held by
participants. On review, the de¿ nitions predominately broke down into two
different categories, those who equated “rurality” with non-urban, agricul-
tural life and those who had no non-urban association with the term at all.
Instead, in this second category, the participants clearly did not have any tra-
ditional understanding of what this term meant whatsoever, and their answers
were often arbitrary. Since some participants had a very different concept of
rurality, the data from those whose de¿ nitions involved a contrast with urban
life/metropolitan lifestyles and those whose de¿ nitions did not ¿ t any tradi-
tional understanding of rurality were split and run as two separate groups, the
traditional de¿ ners and the non-traditional de¿ ners. The results from the two
groups were then compared and the results were also compared to those from
the earlier perception studies.
Table 17.7 shows descending mean score results for the traditional and
non-traditional de¿ ners as well as for the original “Southernness” accuracy
perception test. The comparative runs for traditional vs. non-traditional
de¿ ners within the ruralness test clearly showed that this de¿ nitional dis-
tinction affected participants’ perception and evaluation of vowel variants.